Internal Ballast vs. External Ballast
On a boat with external ballasting in a hard grounding, the grounding tries to drive the keel aft trying to sheer the bolts though the glass at the sump, there is a vertical force that tries to push the keel upwards through the hull, and depending on the length of the ballast keel, a set of large forces that pull downward on the forward bolts and try to drive the aft end of the keel upward through the bottom of the boat. In a properly constructed hull there are athwartships frames that take these loads and distribute them out into the hull. The bolts have large washers that form double sheer for the keel bolts reducing the chance of slicing back through the skin and the bottom of the keel sump is extremely thick resulting from the lapping of the hull fiberglass at the centerline as well as the fiberglass flanges that are formed on the athwartship frames when they are glassed in.
One of the nice things about a cast lead keel vs a cast iron keel is that the physical properties of lead allow it to absorb a lot of the impact without transmitting as much of the loads into the hull. Cast iron on the other hand lacks the ability to absorb loads and so transmits the loads more directly.
In a severe grounding or with a less than perfect construction in this area, there can be damage to the hull laminate near the sump. For example, a friend''s boat that was dropped on its keel when being offloaded from a ship, there was some localized damage to the keel sump but most of the damage was dislocation of the athwartships frames. There was also widely circulated photos of a J-35 that hit the rocks in Maine and lost its keel in which the entire sump structure was torn loose by the impact. (In my mind this set of photos appear to support my contention that coring should stop well short of the keel sump turndown)
To answer the second part of your question, there are a number of issues with regard to voids between the ballast and the ballast keel. To begin with once a void forms it will often permit the condesation of water that has permiated the laminate. Water will also enter these voids through the comparatively thin membrane above the ballast.
In a northern climate this water can and does freeze in winter exerting pressure on the bond between the ballast and the encapsolation. This pressure will enlarge the delamination over time. The presence of water on the interior of the encapsulation is more likely to cause osmotic blistering from the inside out and when you talk about a large percentage of the encapsulation having voids this can represent a significant decrease in the strength of the encapsolation membrane. When dealing with ferrous ballast or worse yet a combination of loose ferrous ballast contained in cement or resin binder, this moisture can cause corrosion. That corrosion causes an expeansion of the surface of the metal which further pries at the bond between the binder and between the ballast and the encapsolation increasing the size of the voids and breaking the bond between the ferrous materials and thier binders. In the worse cases, (like the Bristol 24 that I have mentioned in a prior discussion) the shape of the exterior of the keel can become deformed.
If you never ran aground these voids don''t mean too much. It would be no worse than having the loose internal ballast that was in use well into the early 20th century.
The problem is with impacts. In a hard grounding, the impact generally crushes the encapsolation at point of impact. Then, similar to a grounding with an external keel, the grounding tries to drive the keel aft though the glass at the sump, there is a vertical force that tries to push the keel upwards through the sealer membrane at the top of the keel, and depending on the length of the ballast keel, a set of large forces that pull downward on the forward end of the ballast and tries to drive the aft end of the keel upward through the bottom of the boat. Without keel bolts the adhession between the encapsolation and the ballast serves to distribute these loads into the skin of the boat and then on into the hull. The voids between the ballast keel and the encapsulation reduce the ability to absorb and distrubute these loads over a large area resulting in a larger, more concentrated load against the sealer membrane at the top of the keel. The voids also provide passages for water that enters at the point of impact to flow upward through the encapsulation and sealer membrane into the bilge (which is what happened in the case of my family''s Pearson Vanguard which I believe that I have previously cited).
When the bond between the ballast and the encapsolation is intact, the ballast holds the encapsulation ''in column'' allowing for a further more even distribution of loads, some taken in the encapsolation and some in the ballast itself. In the absense of the bond, the skind is free to buckle and so far greater loads are absorbed in the ballast and therefore exerted on the sealer mambrane in the bilge. What aggrevates this further is that encapsulated keels rarely have the closely spaced athwartships frames that are necessary for an external keel. This means that there is less restraining the top of the ballast. With fewer or no athwartship frames, this means that the comparatively light weight sealer membrane has to absorb the bulk of the impact which this vital membrane is rarely engineered to do.